Various numerical predictions are given for the containment testing of fume hoods with internal baffles of different sizes and at different locations and with a louvered bypass. The numerical predictions were compared with experimental data and a fairly good agreement for the air velocity was found, although there was only qualitative agreement for the concentration of the contaminant. It was found that the level of contaminant leakage at the working aperture depended not only on the size of the recirculating airflow behind the fume hood sash but also on the blockage caused by the baffles inside the fume hood near the exhaust duct. Finally some new suggestions for the design of the fume hood have been proposed. However, before implementing these suggestions it is essential that either experimental testing be performed, or the cost-effective alternative of using numerical predictions for containment testing, to determine accurately the level of contaminant leakage from the working aperture.